I’ve been thinking about something that’s always felt like a missing piece in OmniFocus: seeing my available tasks with planned dates directly in the iOS Calendar app, alongside my appointments and events.
Apps like GoodTask achieve this seamlessly because they use the Reminders database as their backend. Since EventKit ties Reminders and Calendar together, tasks just show up in your calendar view automatically. It’s a really nice unified experience for planning your day.
Now, I understand why OmniFocus uses its own database — the data model is far more sophisticated than what Reminders can support. Projects, nested tasks, defer dates, perspectives… none of that would be possible with Reminders as a backend. That’s a trade-off I’m happy with.
But here’s an idea: what if OmniFocus maintained a lightweight “satellite” sync with Reminders, just for visualization purposes?
The concept would work like this: tasks that are currently available and have a planned date would be mirrored as reminders in a dedicated list (something like “OmniFocus Tasks”). Only a few fields would need to stay in sync — title, planned date, and completion status. Each reminder could include an omnifocus:///task/… link so tapping it takes you straight to the real task. OmniFocus would own the data; the reminder would just be a projection of it.
This way, when I open Calendar and look at my day, I’d see my meetings and my planned tasks together, without leaving the OmniFocus ecosystem or compromising on its power.
An alternative approach — maybe even simpler — would be to sync these tasks as calendar events instead of reminders. Events natively support duration, so estimated duration could map directly to event length. A dedicated calendar called “OmniFocus” could be toggled on or off in Calendar’s view settings. This might actually be cleaner from an implementation standpoint.
I know there’s the existing “Publish to Calendar” feature, but that’s a static ICS feed — it doesn’t update in real-time and doesn’t handle completion status. What I’m describing would be a live, bidirectional (or at least frequently updated) sync.
Would love to hear if others have felt this gap and what the team thinks about feasibility. For me, this would be the final piece in making OmniFocus truly feel like part of the Apple ecosystem rather than a parallel universe.
As far as new features go, I could list dozens before this would even come up. It’s a solution looking for a problem. In OF, you can see your calendars.
There’s no gap that isn’t imagined. But maybe I don’t understand your point here. But whatever it is, it cannot be more useful than many other things the devs could be spending their time on.
I would argue OmniFocus’ calendar integration in Forecast is handy but it’s also not the best calendar UI out there. I would also hope we stopped pretending it is still hard to output features in the age of Claude Code. We should expect Omnigroup to ship your dozen features and more at a pace 10 times that of 10 years ago. No reason for software to stall anymore. And no reason to overlook the value of small quality of life improvements. The OmniFocus team has done their homework. It’s been rock solid for over a decade and it just over went a major rewrite in SwiftUI to make the codebase multi platform so they could ship features faster. The foundations are done. It’s time now to get the small stuff right, the long tail of use cases. Reach for the moon.
If they could hit 1000 features, this would be not make the list. What problem are you trying to solve? It’s a preference not a problem. I can list off the top of my head, dozens of lower hanging more meaningful fruit than whatever this is.
Claude Code lol. Can we keep the memes out of here? Long tail . . . Reach for the moon . . .
Let’s talk like adults, not product managers.
Surprisingly, this is the first time I’ve seen a comment like this on the forum.
For experienced developers, especially ones which know their products well, AI support for coding doesn’t give the boosts you think it does.
Omni work on features which make their product better, that sometimes includes features their users ask for, but Omni have a very clear plan of where they want the product to go. They do listen to users, but that doesn’t mean that they add every single thing asked for. Features add complexity, both in the interface but in the code.
Saying no is often a good thing, not a bad thing.
That’s a fair point. I do believe we’re a few months away from having available tools that generate UI on demand. If that doesn’t change the expectations around what’s reasonably possible in software development, I’m not sure what might. The type of bitter, unwelcoming reaction I’ve met here is something I would have expected 10 years ago. Now it just feels anachronistic trying to gatekeep what is worthy of being developed and what isn’t. I still think planned dates unlock great opportunities for integration with third party calendars, which is only natural for a task manager. More integrations is good UI. The idea of seeing my relevant tasks just where I want them to be feels aesthetically pleasing to me. I’d hope it would resonate, but it’s really fine that it doesn’t. In a sense, though, it might be a generational thing. The inability of this community of embracing the needs of younger people. Zoomers and young millennials need task management software too, you know. Even the product managers. But next time I think of something and it feels sort of nice and clever, I’ll probably just keep it to myself.
When generated apps get that easy and reliable, you can make your own task app that does what you want. For other developers’ apps, when features become cheaper to develop, the developer’s taste and ability to fully bake ideas becomes more important, not less.
I have made many feature requests to Omni and I write a lot of software, including generated code. I’d never try to browbeat them with my knowledge to get my feature greenlit ahead of others.
You could use Claude Code to make your own .ics service today that syncs with your OF tasks and subscribe to it. It’s not a bad idea. Use what you learned from the design decisions to make a more fully considered feature request.
I swear to god I have no clue why my post offended you so much. I’m not trying to cut in line or anything. In the end of the day, the Omni people will do what the Omni people think it’s best. I just had an idea and tried to present it in a way I thought was persuasive to what I thought was an audience of receptive likeminded people. You could judge the post on its terms or you could ignore it if it’s not your cup of tea. I don’t get being offended that it exists. Really, is this an ego thing, a generational thing, a cultural thing? Is it different forms of autism colliding? Do I suck in English as a second language or is it a huge breach of etiquette to suggest features to developers and no one told me before? Been trying to wrap my head around it.
What you should do is contact the Omni team via the address on the app (under settings), and tell them what you’d like. They don’t necessarily see everything on the forum
You clearly don’t work in application development. Feature requests are placed under rather opinionated and forceful scrutiny by all stake holders. You posted your idea in a forum. Other stake holders are weighing in. You don’t like that which is fine. No one is offended or bothered likely.
If you want to spend a scarce resource EVEN WITH AI assisted coding, you gotta make the case and eat some spine when others argue against its relevance.
I wasn’t aware there was a concerted effort to gatekeep productivity software for developers only. Have a nice day.
And that age will last how long before it’s superseded by the next trendy must have? I have yet to see a 100% vibe coded app that will stand against the output of a senior software engineer.
I could never build something like OmniFocus myself as I’m not a senior software engineer and couldn’t match that level of sophistication. But my own experience coding with the help of AI assistants and being able to increase my output by 1 or 2 orders of magnitude has led me to believe code scarcity is not what it used to be. I can only imagine what the impact would be for a real developer. That’s what I meant by “the age of Claude Code” (a figure of speech, BTW): llm agents have changed everything. I might be wrong though. Still, the idea that a forum dedicated to the diverse, international community of USERS of OmniFocus should only validate the points of view of professional developers is utter nonsense.
This is getting a bit off topic but here is AI’s own take on it…
Senior software developers exhibit a mixed level of trust towards AI-generated code. Surveys and analyses reveal that while a significant number of experienced developers use AI tools extensively, their trust in AI-generated code is conditional. Approximately 32% of senior developers generate over half of their shipped code using AI, indicating a higher usage compared to their junior counterparts. However, they often spend considerable time reviewing and correcting AI-generated code to ensure its reliability and security , .
The cautious approach stems from potential security risks, the need for maintainability, and the reliability of complex business logic. Developers often treat AI output as they would external contributions, requiring thorough validation and manual adjustments, especially for critical systems , .
Despite these reservations, many senior developers recognize the time-saving potential of AI, as it assists in generating boilerplate code and completing routine tasks faster. However, senior developers emphasize the importance of human oversight to mitigate risks and ensure that AI-generated solutions integrate smoothly into existing systems , .
Overall, senior developers are adopting AI tools but maintain an essential role in verifying and refining AI-produced code to uphold quality and functionality standards in software development .
In other words, it can’t be trusted… yet
I’d suggest it’s not cautious. It’s a professional and diligent approach.
BTW that’s the AI I quoted, not @RJN
While AI can present a solution that sometimes works, there are no guarantee that the provided code is effective, efficient, or secure.
There are already services available to help with the emerging problem of poor AI code.
